A Continuous Adherence Score, but Not Dichotomous BAASIS Screening, Is Associated with Graft Function in Kidney Transplant Recipients

Background/Objectives: Non-adherence to immunosuppressive medication is a principal modifiable determinant of kidney allograft survival, yet its associations with objective graft function remain inconsistent, in part because the Basel Assessment of Adherence to Immunosuppressive Medications Scale (BAASIS) dichotomises infrequent and persistent non-adherence together. We tested whether a continuous adherence measure and psychological acceptance of illness are more closely associated with graft function than BAASIS. Methods: Cross-sectional, single-centre study of 230 adult kidney transplant recipients; the primary analysis included 199 recipients with eGFR measured on the day of assessment and complete covariate data. Approximately 95% of recipients had received a deceased-donor and 5% a living-donor graft. Adherence was measured by BAASIS, the Adherence in Chronic Diseases Scale (ACDS), and the Adherence to Refills and Medications Scale (ARMS). Psychosocial covariates included the Acceptance of Illness Scale (AIS), Hospital Anxiety and Depression Scale—Modified (HADS-M), WHO Quality of Life—BREF (WHOQOL-BREF), Tilburg Frailty Indicator (TFI), and Mini-Mental State Examination (MMSE). The primary outcome was eGFR on the day of assessment; secondary and exploratory outcomes were serum creatinine and the individual eGFR slope reconstructed from annual archival records (n = 147). Multivariable linear regression adjusted for age, sex, time since transplantation, and psychosocial covariates was complemented by a linear mixed-effects model (158 recipients, 747 eGFR observations) and by sensitivity analyses, including adjustment for the immunosuppressive regimen. Results: Among 199 recipients (46.2% female; mean age 58.5 ± 14.0 years; median time since transplantation 6.5 years; calcineurin inhibitor in 95.2% of those with regimen data), 36 (18.1%) met the BAASIS criterion for non-adherence, most frequently due to timing deviation > 2 h. BAASIS non-adherence was not associated with eGFR (p = 0.14) or eGFR slope (p = 0.23). ACDS correlated with eGFR (ρ = +0.22, p = 0.002) and was independently associated with higher eGFR (β = +1.78 mL/min/1.73 m² per ACDS point, 95% CI +0.56 to +3.00; p = 0.005; adjusted R² = 0.083) and with lower log-creatinine (β = −0.051; p = 0.025). Mean eGFR across ACDS tertiles was 35.9, 40.5, and 44.6 mL/min/1.73 m² (Kruskal–Wallis p = 0.034). The association was robust to adjustment for immunosuppressive regimen and to removal of the least consistent ACDS item. The mixed-effects model showed an average eGFR decline of 1.21 mL/min/1.73 m² per year (p < 0.001), with no significant adherence × time or psychosocial × time interactions. Acceptance of illness was not independently associated with eGFR (p = 0.17). Conclusions: In kidney transplant recipients, dichotomous BAASIS screening was not associated with graft function, whereas higher continuous ACDS scores were independently associated with higher eGFR. Given the cross-sectional design, the low internal consistency of the ACDS in this sample, and the modest explained variance, these findings are hypothesis-generating and warrant prospective confirmation with objective adherence measures.